Padding budget projections to avoid deficits “come hell or high water” is a flawed practice that has outlived its usefulness, says a new C.D. Howe Institute study.

The Commentary, Beating the Odds: A New Framework for Prudent Federal Budgeting, documents how the extra fiscal room built into recent federal budgets has increasingly fuelled in-year spending binges.

The study’s author, William Robson, senior vice president and director of research at the institute, points out that in the last five fiscal years, Ottawa has had some $34 billion more in revenue increases and interest savings than it budgeted. During the same period, spending hikes outran the budget projections by $27 billion, almost $18 billion of which was in 2004/05 alone.

Using up the fiscal room in this way is swelling the size of the federal government, pre-empting debt pay-down and tax relief, and threatens abrupt cuts to spending if revenue flags, argues Robson,.

According to the study, a better approach would formally recognize that surprises will always make Ottawa’s bottom line vary from what was budgeted, and make fiscal projections using a model that takes uncertainties into account. Robson shows that doing multiple runs of this model in which uncertain variables — such as economic growth, tax yields, and interest rates — vary from run to run and gives a more realistic view of the risks associated with a given fiscal plan.

Robson points out several advantages of a probability-based approach to debt targets or avoiding deficits. Establishing a range of “respectable” error around the bottom line would reduce the embarrassment of point forecasts that are inevitably missed, helping Parliament steer a steady course through temporary swings in fortune.

He argues that it would mitigate the pressure to spend first and ask why and how later, evident in recent budgets and in legislation, such as Bill C-48, that explicitly authorizes last-minute splurges.

The study also suggests making a new agency responsible for the probability-based modeling, to avoid the appearance of conflict of interest when the Department of Finance prepares both the projections and the fiscal plan, and to promote open discussion of the uncertainties that bedevil fiscal planning, but are awkward for the minister of finance to acknowledge.

The hell-or-high-water framework, the study concludes, offers Canadians poor odds of achieving a sustainable path to lower debt and taxes. A probability-based approach could improve those odds.